............................... 1
.......................................... 1
.......................... 1
......................... 2
.................................... 5
................................... 6
........................... 2
.......................................... 2
........... 3
..................................... 14
...........................16
......................................... 16
.............. 16
...................................... 18
.............................. 23
.................................. 24
................. 19
..................22
......................................... 22
............................. 22
.................................. 24
............................. 26
....................................... 27
1 :
1-1-
1
2
3
2-1-
1
Parkinson
2
Neurotransmitter
3
Dopamine
2
1
2
3
4
5
6
(
7
.
3-1-
1
Dyskinesia
2
Rigidity
3
Bradykinesia
4
Hypomimia
5
Facial Masking
6
Dimentia
7
Autonomic Nervous System
3
1
2
1-3-1-
3
1987
1
CT Scan
2
MRI
3
Unified Parkinson's Disease Rating Scale(UPDRS)
4
6 1
23
4
1
2
2-3-1-
1
MDS-UPDRS
2
Movement Disorders Society (MDS)
5
٪
8
9
4-1-
6
5-1-
6-1-
2
2 :
1-2-
1
1
Dysarthria
1
2-2-
1-2-2-
1
Micrographia
1
2
3
4
5
6
7
8
1
Geometric features
2
Facial key points
3
Texture features
4
Local Binary Pattern (LBP)
5
Gray-Level Co-Occurrence Matrix (GLCM)
6
Histogram of Oriented Gradients (HOG)
7
Facial Action Coding System (FACS)
8
Action Unit
1
2
3
4
5
6
1
Augmented
2
Convolutional Neural Network (CNN)
3
Bayesian Neural Network
4
Residual Neural Network
5
Palpebral Fissure
6
Logistic Regression
%
1
2
3
4
٪
1
Jakubowski
2
Fusion
3
AlexNet
4
XGBoost
1
2
3
4
5
6
8
7
7
8
9
1
SVM
2
Face mobility index
3
KNN
4
Random Forest
5
Transfer learning
6
Triplet loss
7
Visual Geometric Graph (VGG)
8
ResNet
9
Multiframe
1
2-2-2-
2
1
YouTube
2
Hypophonia
1
2
3
4
5
6
7
1
Dysarthria
2
Hypokinetic dysarthria
3
Static
4
Dynamic
5
Jitter
6
Shimmer
7
MFCCs
1
.
2
3
4
5
1
HNR
2
Formants
3
Pitch contour
4
Stop consonant
5
MFCC
1
32
33
3134
2
3-2-2-
1
Principal Component Analysis (PCA)
2
Spectrogram
1
.
2
1
Micrographia
2
Supervised learning
37
1
27
8580
76 11
2
96٪
38
99.22
4-2-2-
3
4
1
Bernardo
2
Naïve Bayes
3
Multimodal
4
Joshi
1
2
3-2-
1
Postural tremor
2
Rest tremor
16
3 :
1-3-
2-3-
17
3-1
)(
)ج
أ
18
21
3-1
3-3-
-
1
1
Scream go hero
3-2
19
4-3-
1
2
3
1
Interpretable
2
Confidence level
3
Margin of error
20
1
2
3-3
1
Risk factors
2
Scalability
21
1
2
3-3
1
Weak Supervision
2
Bayesian Networks
22
4 :
1-4-
2-4-
23
1
3-4-
25
25
1.
2.
3.
4.
5.
1
Usability
24
1.
2.
3.
4-4-
5-4-
1
2
1
Sensitivity
2
Specificity
25
1
2
1
Positive Predictive Value(PPV)
2
Likelihood Ratios (LR)+
26
1
1
2
3
6-4-
1
Pearson correlation coefficient
2
Spearman's rank correlation coefficient
3
Intraclass crrelation coefficient
27
1
7-4-
28
1
29
5 :
[1] S. J. Chinta and J. K. Andersen, "Redox imbalance in Parkinson's disease,"
Biochimica et Biophysica Acta (BBA)-General Subjects, vol. 1780, no. 11, pp.
1362-1367, 2008.
[2] C. G. Goetz et al., "Movement Disorder Society‐sponsored revision of the Unified
Parkinson's Disease Rating Scale (MDS‐UPDRS): scale presentation and
clinimetric testing results," Movement disorders: official journal of the Movement
Disorder Society, vol. 23, no. 15, pp. 2129-2170, 2008.
[3] E. Fitzpatrick, N. Hohl, P. Silburn, C .O’Gorman, and S. A. Broadley, "Case–
control study of blink rate in Parkinson’s disease under different conditions,"
Journal of neurology, vol. 259, no. 4, pp. 739-744, 2012.
[4] H. M. Gray and L. Tickle-Degnen, "A meta-analysis of performance on emotion
recognition tasks in Parkinson’s disease," Neuropsychology, vol. 24, no. 2, p. 176,
2010.
[5] F. L. Pagan, "Improving outcomes through early diagnosis of Parkinson's
disease," American Journal of Managed Care, vol. 18, no. 7, p. S176, 2012.
[6] S. Fahn, "Members of the UPDRS development committee. Unified Parkinson's
disease rating scale," Recent developments in Parkinson's disease, vol. 2, pp. 293-
304, 1987.
[7] M. D. Pell and C. L. Leonard, "Facial expression decoding in early Parkinson's
disease," Cognitive Brain Research, vol. 23, no. 2-3, pp. 327-340, 2005.
30
[8] G. Defazio, M. Guerrieri, D. Liuzzi, A. F. Gigante, and V. Di Nicola, "Assessment
of voice and speech symptoms in early Parkinson’s disease by the Robertson
dysarthria profile," Neurological Sciences, vol. 37, pp. 443-449, 2016.
[9] J. M. Tracy, Y. Özkanca, D. C. Atkins, and R. H. Ghomi, "Investigating voice as
a biomarker: deep phenotyping methods for early detection of Parkinson's
disease," Journal of Biomedical Informatics, vol. 104, p. 1033 62 ,2020.
[10] D. Ghimire and J. Lee, "Geometric feature-based facial expression recognition in
image sequences using multi-class adaboost and support vector machines,"
Sensors, vol. 13, no. 6, pp. 7714-7734, 2013.
[11] T. Ahonen, A. Hadid, and M. Pietikäinen, "Face recognition with local binary
patterns," in European conference on computer vision, 2004: Springer, pp. 469-
481 .
[12] A. Baraldi and F. Panniggiani, "An investigation of the textural characteristics
associated with gray level cooccurrence matrix statistical parameters," IEEE
transactions on geoscience and remote sensing, vol. 33, no. 2, pp. 293-304, 1995.
[13] N. Dalal and B. Triggs, "Histograms of oriented gradients for human detection,"
in 2005 IEEE computer society conference on computer vision and pattern
recognition (CVPR'05), 2005, vol. 1: Ieee, pp. 886-893 .
[14] E. Friesen and P. Ekman, "Facial action coding system: a technique for the
measurement of facial movement," Palo Alto, vol. 3, no. 2, p. 5, 1978.
[15] P. Ekman, W. V. Friesen ,and J. C. Hager, Facial Action Coding System: Facial
action coding system: the manual: on CD-ROM. Research Nexus, 2002.
[16] A. Grammatikopoulou, N. Grammalidis, S. Bostantjopoulou, and Z. Katsarou,
"Detecting hypomimia symptoms by selfie photo analysis :for early Parkinson
disease detection," in Proceedings of the 12th ACM International Conference on
PErvasive Technologies Related to Assistive Environments, 2019, pp. 517-522 .
[17] J. Jakubowski, A. Potulska-Chromik, K. Białek, M. Nojszewska, and A. Kostera-
Pruszczyk, "A Study on the Possible Diagnosis of Parkinson’s Disease on the
Basis of Facial Image Analysis," Electronics, vol. 10, no. 22, p. 2832, 2021.
[18] M. Rajnoha, J. Mekyska, R. Burget, I. Eliasova, M. Kostalova, and I. Rektorová,
"Towards identification of hypomimia in Parkinson's disease based on face
recognition methods," in 2018 10th International Congress on Ultra Modern
Telecommunications and Control Systems and Workshops (ICUMT), 2018: IEEE,
pp. 1-4 .
[19] B. Sonawane and P. Sharma, "Review of automated emotion-based quantification
of facial expression in Parkinson’s patients," The Visual Computer, vol. 37, no. 5,
pp. 1151-1167, 2021.
[20] N. Vinokurov, D. Arkadir, E. Linetsky, H. Bergman, and D. Weinshall,
"Quantifying hypomimia in parkinson patients using a depth camera," in
Pervasive Computing Paradigms for Mental Health: 5th International
Conference, MindCare 2015, Milan, Italy, September 24-25, 2015, Revised
Selected Papers 5, 2016: Springer, pp. 63-71 .
[21] J. Skibińska and R. Burget, "Parkinson’s disease detection based on changes of
emotions during speech," in 2020 12th International Congress on Ultra Modern
Telecommunications and Control Systems and Workshops (ICUMT), 2020: IEEE,
pp. 124-130 .
31
[22] M. R. Ali, T. Myers, E. Wagner, H. Ratnu, E. Dorsey, and E. Hoque, "Facial
expressions can detect Parkinson’s disease: Preliminary evidence from videos
collected online," NPJ digital medicine, vol. 4, no. 1, pp. 1-4, 2021.
[23] Y. Guan, "Application of logistic regression algorithm in the diagnosis of
expression disorder in Parkinson's disease," in 2021 IEEE 2nd International
Conference on Information Technology, Big Data and Artificial Intelligence
(ICIBA), 2021, vol. 2: IEEE, pp. 1117-1120 .
[24] E. Pegolo, D. Volpe, A. Cucca, L .Ricciardi, and Z. Sawacha, "Quantitative
Evaluation of Hypomimia in Parkinson’s Disease: A Face Tracking Approach,"
Sensors, vol. 22, no. 4, p. 1358, 2022.
[25] L. F. Gomez, A. Morales, J. Fierrez, and J. R. Orozco-Arroyave, "Exploring facial
expressions and action unit domains for Parkinson detection," Plos one, vol. 18,
no. 2, p. e0281248, 2023.
[26] A. Abrami, S. Gunzler, C. Kilbane, R. Ostrand, B. Ho, and G. Cecchi, "Automated
computer vision assessment of hypomimia in Parkinson disease: proof-of-
principle pilot study," Journal of Medical Internet Research, vol. 23, no. 2, p.
e21037, 2021.
[27] A. Garcia-Agundez, T. Ochs, R. Konrad, P. Caserman, and S. Göbel, "Eye Aspect
Ratio Based Blink Rate Detection and Its Potential Use for Parkinson’s Disease ",
2018.
[28] A. Shinde, R. Atre, A. S. Guleria, R. Nibandhe, and R. Shriram, "Facial Features
Based Prediction of Parkinson's Disease," in 2018 3rd International Conference
for Convergence in Technology (I2CT), 2018: IEEE, pp. 1-5 .
[29] H. Iwaki et al ." , Using Spontaneous Eye-blink Rates to Predict the Motor Status
of Patients with Parkinson's Disease," Internal Medicine, vol. 58, no. 10, pp. 1417-
1421, 2019.
[30] S. Sapir, L. Ramig, and C. Fox, "Speech and swallowing disorders in Parkinson
disease," Current opinion in otolaryngology & head and neck surgery, vol. 16,
no. 3, pp. 205-210, 2008.
[31] W. Rahman et al., "Detecting Parkinson disease using a web-based speech task:
Observational study," Journal of Medical Internet Research, vol. 23, no. 10, p. e2,
2021.
[32] Z. Ayaz, S. Naz, N. H. Khan, I. Razzak, and M. Imran, "Automated methods for
diagnosis of Parkinson’s disease and predicting severity level," Neural Computing
and Applications, pp. 1-36, 2022.
[33] R. Mathur, V. Pathak, and D. Bandil, "Parkinson disease prediction using machine
learning algorithm," in Emerging Trends in Expert Applications and Security:
Proceedings of ICETEAS 2018, 2019: Springer, pp. 357-363 .
[34] L. Parisi, N. RaviChandran, and M. L. Manaog, "Feature-driven machine learning
to improve early diagnosis of Parkinson's disease," Expert Systems with
Applications, vol. 110, pp. 182-190, 2018.
[35] L. Zahid et al., "A spectrogram-based deep feature assisted computer-aided
diagnostic system for Parkinson’s disease," IEEE Access, vol. 8, pp. 35482-35495,
2020.
[36] E. Heremans, E. Nackaerts, S. Broeder, G. Vervoort, S. P. Swinnen, and A.
Nieuwboer, "Handwriting impairments in people with Parkinson’s disease and
32
freezing of gait," Neurorehabilitation and neural repair, vol. 30 ,no. 10, pp. 911-
919, 2016.
[37] M. Gil-Martín, J. M. Montero, and R. San-Segundo, "Parkinson’s disease
detection from drawing movements using convolutional neural networks,"
Electronics, vol. 8, no. 8, p. 907, 2019.
[38] I. Kamran, S. Naz, I. Razzak, and M. Imran, "Handwriting dynamics assessment
using deep neural network for early identification of Parkinson’s disease," Future
Generation Computer Systems, vol. 117, pp. 234-244, 2021.
[39] A. Joshi, S. Ghosh, S. Gunnery, L. Tickle-Degnen, S. Sclaroff, and M. Betke,
"Context-sensitive prediction of facial expressivity using multimodal hierarchical
bayesian neural networks," in 2018 13th IEEE International Conference on
Automatic Face & Gesture Recognition (FG 2018), 2018: IEEE, pp. 278-285 .
[40] S. Arora et al., "Smartphone motor testing to distinguish idiopathic REM sleep
behavior disorder, controls, and PD," Neurology, vol. 91, no. 16, pp. e1528-e1538,
2018.
[41] W. Chango, J. A. Lara, R. Cerezo, and C. Romero, "A review on data fusion in
multimodal learning analytics and educational data mining," Wiley
Interdisciplinary Reviews: Data Mining and Knowledge Discovery, vol. 12, no. 4,
p. e1458, 2022.
33
6 :
Action Unit
Augmented
Autonomic Nervous System
Bayesian Network
Bayesian Neural Network
Bradykinesia
CT Scan
Confidence level
Convolutional Neural Network (CNN)
Dimentia
Dopamine
Dynamic
Dysarthria
Dyskinesia
Facial Action Coding System (FACS)
Facial Masking
Facial Key Points
Formants
34
Fusion
Geometric features
Histogram of Oriented Gradients
(HOG)
Hypokinetic dysarthria
Hypomimia
Hypophonia
Interpretable
Jitter
KNN
Local Binary Pattern (LBP)
Logistic Regression
Margin of error
Micrographia
Movement Disorders Society (MDS)
Multimodal
Naïve Bayes
Neurotransmitter
Palpebral Fissure
Parkinson
Pitch contour
Postural tremor
Principal Component Analysis (PCA)
Random Forest
ResNet
Residual Neural Network
35
Rest tremor
Rigidity
Risk factors
SVM
Scalability
Sensitivity
Shimmer
Specificity
Spectrogram
Static
Supervised learning
Texture features
Transfer learning
Triplet loss
Usability
Weak Supervision
36
Abstract:
Parkinson's disease is a progressive neurological disorder characterized by both motor
and non-motor symptoms. The motor symptoms of this disease manifest as reduced
speed in movement, stiffness, rigidity, and tremors. These symptoms can affect various
body parts and muscles, including the face, which could potentially impede the natural
expression of facial emotions. Parkinson's patients often also experience speech
disorders such as dysarthria. Diagnosis of the disease is crucial because treatment helps
manage the progression of the disease. In recent years, numerous efforts have been
made to develop efficient diagnostic methods for Parkinson's assessment. These
methods are based on various Parkinson's symptoms and utilize different data sources
such as speech, movement patterns, radiological images, handwriting samples, and
videos
In this research study, our goal is to introduce a comprehensive method for the
assessment of individuals with the disease correlated with their MDS-UPDRS test. This
method employs a mobile application that collects visual, voice, handwriting, and other
clinical symptom data from the individual. It then assesses the severity of the disease in
the individual based on all this information. This method not only assists people in self-
assessment before visiting a doctor but also helps specialists in more precise diagnosis
and symptom examination. One of the innovative aspects of this study is the combined
consideration of voice, image, and other modalities. Additionally, given the significance
of accuracy and reliabilityin in medical applications, our proposed method is based on
diagnosing each of the disease's symptoms and providing a causal and explainable result
for the disease. The aim is to create a PD detection model using weak supervision and
Bayesian Networks. In developing this method, we will account for model robustness in
the face of uncertainty, limited, noisy, and incomplete data
The model relies on weak supervision classifiers, which are capable of learning from
limited or noisy supervision. The weak supervision classifiers leverage the LFs defined to
generate probabilistic labels for the data instances. For each symptom such as
hypokinetic dysarthria, hypomimia, and micrographia, we create separate weak
supervision classifiers ensuring that each classifier is optimized for its respective data
source. The defined LFs combine outputs into a single probabilistic label for each data
instance using the learned generative model. These probabilistic labels can then train a
Denoising Network (DN) such as a deep neural network to make predictions on new,
unseen data. The loss function will be defined between the output of the DN and the
labels generated by the generative model. In this case, a Bayesian Network (BN) was
utilized as a fusion method to reach a final decision in a probabilistic framework. By
integrating the outputs of weak supervision models and risk factors, the BN can provide
a more accurate and reliable diagnosis, taking into account prior knowledge. Using BN,
it addresses interpretability and uncertainty quantification, two of the most critical
factors in medical decision-making.
